Nvidia Makes AI computing possible in Cameras
Why You Should Work There
Nvidia is a company most don’t associate with cutting edge technology. Most just think of them as a “graphics board” company.
Into “edge computing” and AI, Nvidia is impressing the world with its continuous technology innovations.
If you’re in the technology field in any significant way, then Nvidia should be a top consideration for you. Unless you’ve kept up with them recently you might still automatically think gaming when you hear that name, but their newest innovations are putting them at the forefront of two very forward-thinking technologies – AI and so-called “edge computing”.
At Media Jobs we are here to present the great companies first
Nvidia is on the Cutting Edge of the “Edge”
If you’re unfamiliar with what is called “the edge”, it’s not difficult to understand (and no, it’s not the guitarist for U2). The data from our devices is pretty broadly sent into the cloud today for storage and processing. That’s unsustainable going forward, according to many people, for several reasons. The primary two are probably bandwidth and latency, with privacy being a close third place. Privacy is a common sense reason – the less data is moved around and transmitted, the less chance there is for intrusion or being hacked.
Bandwidth limitations come from the increasing amount of video being uploaded and the increasingly high definition nature of that video. As more 4K video is used and uploaded, the less bandwidth will be available. Latency is also a big concern for the future as we get closer to technologies in the realm of self-driving cars, which need to process information and make decision in a split second. If it takes your car a few seconds to decide that the wreck in front of you on the highway needs to be avoided, well… that’s not really ideal, is it?
The answer is to have more data processing done on the edge – not in the cloud but at the point of collection, such as right inside the camera or drone doing the collecting. That’s where Nvidia is focusing much of their efforts now, and the results are impressive.
Meet George Jetson, v2.0
Nvidia’s Jetson TX1 card for embedded computing was impressive. Their new Jetson TX2 card is even more amazing, allowing for twice the processing speed of the TX1 or, if that’s unnecessary for the application in question, half of the power required with the TX1 for the same processing power. Lots of companies have already begun using the Jetson TX2 in upcoming projects, and the applications are pretty cool to say the least.
For instance, Live Planet introduced a 4K, 360-degree 3D camera that uses AI to encode 3D video in real time using Jetson. The camera produces a stream of 65 GB, which is just too much to upload to a cloud server. Teal Drones will start shipping a smart drone in a few months, using Jetson TX2 cards, that will be able to understand and react to what it’s cameras are seeing in real time. The implications are both exciting and a little scary.
Investment banks are telling people to buy Nvidia stock and predicting that they’ll hurt the competition, like AMD. We just think that this is a great opportunity to get on board with a tech company that isn’t resting on their laurels and may become one of the major players in forging our high tech future. What do you think?